Microsoft Executive Says SaaS Is Becoming AI Agent Infrastructure
Bryan Goode says AI agents may coordinate work across enterprise applications while those systems retain their data, business logic and governance.
Microsoft Corporate Vice President Bryan Goode said software-as-a-service is becoming infrastructure for artificial intelligence agents, which could change how employees work across enterprise applications.
In a commentary published Sept. 26, 2026, at 8:30 a.m. ET (12:30 p.m. UTC), Goode wrote that the key question is what enterprise applications become as agents handle more work between them.
“I think the better question is what the app becomes,” Goode wrote.
Agents could reduce the manual effort involved in moving information between customer-management, finance, project-management and productivity systems. The applications would continue to hold the underlying data, business rules and governance, while agents coordinate tasks across them.
Goode described the time employees spend moving between systems as the “toggle tax.”
“I call this the ‘toggle tax,’” Goode wrote.
A study involving 137 users across 20 teams at three Fortune 500 companies found that workers switched between applications and websites nearly 1,200 times per day. They spent just under four hours each week reorienting themselves between systems.
Organizational alignment and workflow redesign remain uneven. The 2026 Work Trend Index surveyed 20,000 workers across 10 markets from Feb. 18 to April 7, 2026. Only 26% said their leadership was clearly and consistently aligned on artificial intelligence, while 65% of AI users feared falling behind without faster adoption and 45% said it felt safer to focus on current goals than redesign work around AI.
A 2025 global survey of 1,993 participants found that 39% of respondents said their organizations had begun experimenting with AI agents. Another 23% said they were scaling an agentic AI system in at least one business function.
Separate figures showed 62% of organizations experimenting with AI agents and 39% reporting an enterprise-level earnings impact. Those figures come from separate research and are not directly comparable.
Forecasts point to broader integration of agents into enterprise software. One projection puts agentic AI in 33% of enterprise software applications by 2028, up from less than 1% in 2024. A separate projection says more than 40% of agentic AI projects could be canceled by the end of 2027 because of rising costs, unclear business value or inadequate risk controls.
The approach fits with Microsoft's expansion of Copilot into a workplace AI platform, combining conversational tools, agents and workplace applications under subscription and usage-based billing.
Goode’s view leaves SaaS applications in place but changes how employees interact with them. People may increasingly direct agents, while applications retain control of the information and processes behind the work.
